1 from __future__ import unicode_literals
13 import xml.etree.ElementTree
15 from ..compat import (
20 compat_urllib_parse_urlparse,
40 class InfoExtractor(object):
41 """Information Extractor class.
43 Information extractors are the classes that, given a URL, extract
44 information about the video (or videos) the URL refers to. This
45 information includes the real video URL, the video title, author and
46 others. The information is stored in a dictionary which is then
47 passed to the YoutubeDL. The YoutubeDL processes this
48 information possibly downloading the video to the file system, among
49 other possible outcomes.
51 The type field determines the type of the result.
52 By far the most common value (and the default if _type is missing) is
53 "video", which indicates a single video.
55 For a video, the dictionaries must include the following fields:
58 title: Video title, unescaped.
60 Additionally, it must contain either a formats entry or a url one:
62 formats: A list of dictionaries for each format available, ordered
63 from worst to best quality.
66 * url Mandatory. The URL of the video file
67 * ext Will be calculated from url if missing
68 * format A human-readable description of the format
69 ("mp4 container with h264/opus").
70 Calculated from the format_id, width, height.
71 and format_note fields if missing.
72 * format_id A short description of the format
73 ("mp4_h264_opus" or "19").
74 Technically optional, but strongly recommended.
75 * format_note Additional info about the format
76 ("3D" or "DASH video")
77 * width Width of the video, if known
78 * height Height of the video, if known
79 * resolution Textual description of width and height
80 * tbr Average bitrate of audio and video in KBit/s
81 * abr Average audio bitrate in KBit/s
82 * acodec Name of the audio codec in use
83 * asr Audio sampling rate in Hertz
84 * vbr Average video bitrate in KBit/s
86 * vcodec Name of the video codec in use
87 * container Name of the container format
88 * filesize The number of bytes, if known in advance
89 * filesize_approx An estimate for the number of bytes
90 * player_url SWF Player URL (used for rtmpdump).
91 * protocol The protocol that will be used for the actual
93 "http", "https", "rtsp", "rtmp", "rtmpe",
94 "m3u8", or "m3u8_native".
95 * preference Order number of this format. If this field is
96 present and not None, the formats get sorted
97 by this field, regardless of all other values.
98 -1 for default (order by other properties),
99 -2 or smaller for less than default.
100 < -1000 to hide the format (if there is
101 another one which is strictly better)
102 * language_preference Is this in the correct requested
104 10 if it's what the URL is about,
105 -1 for default (don't know),
106 -10 otherwise, other values reserved for now.
107 * quality Order number of the video quality of this
108 format, irrespective of the file format.
109 -1 for default (order by other properties),
110 -2 or smaller for less than default.
111 * source_preference Order number for this video source
112 (quality takes higher priority)
113 -1 for default (order by other properties),
114 -2 or smaller for less than default.
115 * http_headers A dictionary of additional HTTP headers
116 to add to the request.
117 * stretched_ratio If given and not 1, indicates that the
118 video's pixels are not square.
119 width : height ratio as float.
120 * no_resume The server does not support resuming the
121 (HTTP or RTMP) download. Boolean.
123 url: Final video URL.
124 ext: Video filename extension.
125 format: The video format, defaults to ext (used for --get-format)
126 player_url: SWF Player URL (used for rtmpdump).
128 The following fields are optional:
130 alt_title: A secondary title of the video.
131 display_id An alternative identifier for the video, not necessarily
132 unique, but available before title. Typically, id is
133 something like "4234987", title "Dancing naked mole rats",
134 and display_id "dancing-naked-mole-rats"
135 thumbnails: A list of dictionaries, with the following entries:
136 * "id" (optional, string) - Thumbnail format ID
138 * "preference" (optional, int) - quality of the image
139 * "width" (optional, int)
140 * "height" (optional, int)
141 * "resolution" (optional, string "{width}x{height"},
143 thumbnail: Full URL to a video thumbnail image.
144 description: Full video description.
145 uploader: Full name of the video uploader.
146 creator: The main artist who created the video.
147 timestamp: UNIX timestamp of the moment the video became available.
148 upload_date: Video upload date (YYYYMMDD).
149 If not explicitly set, calculated from timestamp.
150 uploader_id: Nickname or id of the video uploader.
151 location: Physical location where the video was filmed.
152 subtitles: The available subtitles as a dictionary in the format
153 {language: subformats}. "subformats" is a list sorted from
154 lower to higher preference, each element is a dictionary
155 with the "ext" entry and one of:
156 * "data": The subtitles file contents
157 * "url": A url pointing to the subtitles file
158 automatic_captions: Like 'subtitles', used by the YoutubeIE for
159 automatically generated captions
160 duration: Length of the video in seconds, as an integer.
161 view_count: How many users have watched the video on the platform.
162 like_count: Number of positive ratings of the video
163 dislike_count: Number of negative ratings of the video
164 average_rating: Average rating give by users, the scale used depends on the webpage
165 comment_count: Number of comments on the video
166 comments: A list of comments, each with one or more of the following
167 properties (all but one of text or html optional):
168 * "author" - human-readable name of the comment author
169 * "author_id" - user ID of the comment author
171 * "html" - Comment as HTML
172 * "text" - Plain text of the comment
173 * "timestamp" - UNIX timestamp of comment
174 * "parent" - ID of the comment this one is replying to.
175 Set to "root" to indicate that this is a
176 comment to the original video.
177 age_limit: Age restriction for the video, as an integer (years)
178 webpage_url: The url to the video webpage, if given to youtube-dl it
179 should allow to get the same result again. (It will be set
180 by YoutubeDL if it's missing)
181 categories: A list of categories that the video falls in, for example
183 is_live: True, False, or None (=unknown). Whether this video is a
184 live stream that goes on instead of a fixed-length video.
186 Unless mentioned otherwise, the fields should be Unicode strings.
188 Unless mentioned otherwise, None is equivalent to absence of information.
191 _type "playlist" indicates multiple videos.
192 There must be a key "entries", which is a list, an iterable, or a PagedList
193 object, each element of which is a valid dictionary by this specification.
195 Additionally, playlists can have "title" and "id" attributes with the same
196 semantics as videos (see above).
199 _type "multi_video" indicates that there are multiple videos that
200 form a single show, for examples multiple acts of an opera or TV episode.
201 It must have an entries key like a playlist and contain all the keys
202 required for a video at the same time.
205 _type "url" indicates that the video must be extracted from another
206 location, possibly by a different extractor. Its only required key is:
207 "url" - the next URL to extract.
208 The key "ie_key" can be set to the class name (minus the trailing "IE",
209 e.g. "Youtube") if the extractor class is known in advance.
210 Additionally, the dictionary may have any properties of the resolved entity
211 known in advance, for example "title" if the title of the referred video is
215 _type "url_transparent" entities have the same specification as "url", but
216 indicate that the given additional information is more precise than the one
217 associated with the resolved URL.
218 This is useful when a site employs a video service that hosts the video and
219 its technical metadata, but that video service does not embed a useful
220 title, description etc.
223 Subclasses of this one should re-define the _real_initialize() and
224 _real_extract() methods and define a _VALID_URL regexp.
225 Probably, they should also be added to the list of extractors.
227 Finally, the _WORKING attribute should be set to False for broken IEs
228 in order to warn the users and skip the tests.
235 def __init__(self, downloader=None):
236 """Constructor. Receives an optional downloader."""
238 self.set_downloader(downloader)
241 def suitable(cls, url):
242 """Receives a URL and returns True if suitable for this IE."""
244 # This does not use has/getattr intentionally - we want to know whether
245 # we have cached the regexp for *this* class, whereas getattr would also
246 # match the superclass
247 if '_VALID_URL_RE' not in cls.__dict__:
248 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
249 return cls._VALID_URL_RE.match(url) is not None
252 def _match_id(cls, url):
253 if '_VALID_URL_RE' not in cls.__dict__:
254 cls._VALID_URL_RE = re.compile(cls._VALID_URL)
255 m = cls._VALID_URL_RE.match(url)
261 """Getter method for _WORKING."""
264 def initialize(self):
265 """Initializes an instance (authentication, etc)."""
267 self._real_initialize()
270 def extract(self, url):
271 """Extracts URL information and returns it in list of dicts."""
274 return self._real_extract(url)
275 except ExtractorError:
277 except compat_http_client.IncompleteRead as e:
278 raise ExtractorError('A network error has occured.', cause=e, expected=True)
279 except (KeyError, StopIteration) as e:
280 raise ExtractorError('An extractor error has occured.', cause=e)
282 def set_downloader(self, downloader):
283 """Sets the downloader for this IE."""
284 self._downloader = downloader
286 def _real_initialize(self):
287 """Real initialization process. Redefine in subclasses."""
290 def _real_extract(self, url):
291 """Real extraction process. Redefine in subclasses."""
296 """A string for getting the InfoExtractor with get_info_extractor"""
297 return cls.__name__[:-2]
301 return type(self).__name__[:-2]
303 def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
304 """ Returns the response handle """
306 self.report_download_webpage(video_id)
307 elif note is not False:
309 self.to_screen('%s' % (note,))
311 self.to_screen('%s: %s' % (video_id, note))
313 return self._downloader.urlopen(url_or_request)
314 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
318 errnote = 'Unable to download webpage'
319 errmsg = '%s: %s' % (errnote, compat_str(err))
321 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
323 self._downloader.report_warning(errmsg)
326 def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None):
327 """ Returns a tuple (page content as string, URL handle) """
328 # Strip hashes from the URL (#1038)
329 if isinstance(url_or_request, (compat_str, str)):
330 url_or_request = url_or_request.partition('#')[0]
332 urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
336 content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding)
337 return (content, urlh)
340 def _guess_encoding_from_content(content_type, webpage_bytes):
341 m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
343 encoding = m.group(1)
345 m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
346 webpage_bytes[:1024])
348 encoding = m.group(1).decode('ascii')
349 elif webpage_bytes.startswith(b'\xff\xfe'):
356 def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
357 content_type = urlh.headers.get('Content-Type', '')
358 webpage_bytes = urlh.read()
359 if prefix is not None:
360 webpage_bytes = prefix + webpage_bytes
362 encoding = self._guess_encoding_from_content(content_type, webpage_bytes)
363 if self._downloader.params.get('dump_intermediate_pages', False):
365 url = url_or_request.get_full_url()
366 except AttributeError:
368 self.to_screen('Dumping request to ' + url)
369 dump = base64.b64encode(webpage_bytes).decode('ascii')
370 self._downloader.to_screen(dump)
371 if self._downloader.params.get('write_pages', False):
373 url = url_or_request.get_full_url()
374 except AttributeError:
376 basen = '%s_%s' % (video_id, url)
378 h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
379 basen = basen[:240 - len(h)] + h
380 raw_filename = basen + '.dump'
381 filename = sanitize_filename(raw_filename, restricted=True)
382 self.to_screen('Saving request to ' + filename)
383 # Working around MAX_PATH limitation on Windows (see
384 # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
386 absfilepath = os.path.abspath(filename)
387 if len(absfilepath) > 259:
388 filename = '\\\\?\\' + absfilepath
389 with open(filename, 'wb') as outf:
390 outf.write(webpage_bytes)
393 content = webpage_bytes.decode(encoding, 'replace')
395 content = webpage_bytes.decode('utf-8', 'replace')
397 if ('<title>Access to this site is blocked</title>' in content and
398 'Websense' in content[:512]):
399 msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
400 blocked_iframe = self._html_search_regex(
401 r'<iframe src="([^"]+)"', content,
402 'Websense information URL', default=None)
404 msg += ' Visit %s for more details' % blocked_iframe
405 raise ExtractorError(msg, expected=True)
406 if '<title>The URL you requested has been blocked</title>' in content[:512]:
408 'Access to this webpage has been blocked by Indian censorship. '
409 'Use a VPN or proxy server (with --proxy) to route around it.')
410 block_msg = self._html_search_regex(
411 r'</h1><p>(.*?)</p>',
412 content, 'block message', default=None)
414 msg += ' (Message: "%s")' % block_msg.replace('\n', ' ')
415 raise ExtractorError(msg, expected=True)
419 def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5, encoding=None):
420 """ Returns the data of the page as a string """
423 while success is False:
425 res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal, encoding=encoding)
427 except compat_http_client.IncompleteRead as e:
429 if try_count >= tries:
431 self._sleep(timeout, video_id)
438 def _download_xml(self, url_or_request, video_id,
439 note='Downloading XML', errnote='Unable to download XML',
440 transform_source=None, fatal=True, encoding=None):
441 """Return the xml as an xml.etree.ElementTree.Element"""
442 xml_string = self._download_webpage(
443 url_or_request, video_id, note, errnote, fatal=fatal, encoding=encoding)
444 if xml_string is False:
447 xml_string = transform_source(xml_string)
448 return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
450 def _download_json(self, url_or_request, video_id,
451 note='Downloading JSON metadata',
452 errnote='Unable to download JSON metadata',
453 transform_source=None,
454 fatal=True, encoding=None):
455 json_string = self._download_webpage(
456 url_or_request, video_id, note, errnote, fatal=fatal,
458 if (not fatal) and json_string is False:
460 return self._parse_json(
461 json_string, video_id, transform_source=transform_source, fatal=fatal)
463 def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
465 json_string = transform_source(json_string)
467 return json.loads(json_string)
468 except ValueError as ve:
469 errmsg = '%s: Failed to parse JSON ' % video_id
471 raise ExtractorError(errmsg, cause=ve)
473 self.report_warning(errmsg + str(ve))
475 def report_warning(self, msg, video_id=None):
476 idstr = '' if video_id is None else '%s: ' % video_id
477 self._downloader.report_warning(
478 '[%s] %s%s' % (self.IE_NAME, idstr, msg))
480 def to_screen(self, msg):
481 """Print msg to screen, prefixing it with '[ie_name]'"""
482 self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
484 def report_extraction(self, id_or_name):
485 """Report information extraction."""
486 self.to_screen('%s: Extracting information' % id_or_name)
488 def report_download_webpage(self, video_id):
489 """Report webpage download."""
490 self.to_screen('%s: Downloading webpage' % video_id)
492 def report_age_confirmation(self):
493 """Report attempt to confirm age."""
494 self.to_screen('Confirming age')
496 def report_login(self):
497 """Report attempt to log in."""
498 self.to_screen('Logging in')
500 # Methods for following #608
502 def url_result(url, ie=None, video_id=None, video_title=None):
503 """Returns a url that points to a page that should be processed"""
504 # TODO: ie should be the class used for getting the info
505 video_info = {'_type': 'url',
508 if video_id is not None:
509 video_info['id'] = video_id
510 if video_title is not None:
511 video_info['title'] = video_title
515 def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
516 """Returns a playlist"""
517 video_info = {'_type': 'playlist',
520 video_info['id'] = playlist_id
522 video_info['title'] = playlist_title
523 if playlist_description:
524 video_info['description'] = playlist_description
527 def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
529 Perform a regex search on the given string, using a single or a list of
530 patterns returning the first matching group.
531 In case of failure return a default value or raise a WARNING or a
532 RegexNotFoundError, depending on fatal, specifying the field name.
534 if isinstance(pattern, (str, compat_str, compiled_regex_type)):
535 mobj = re.search(pattern, string, flags)
538 mobj = re.search(p, string, flags)
542 if not self._downloader.params.get('no_color') and os.name != 'nt' and sys.stderr.isatty():
543 _name = '\033[0;34m%s\033[0m' % name
549 # return the first matching group
550 return next(g for g in mobj.groups() if g is not None)
552 return mobj.group(group)
553 elif default is not NO_DEFAULT:
556 raise RegexNotFoundError('Unable to extract %s' % _name)
558 self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message())
561 def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
563 Like _search_regex, but strips HTML tags and unescapes entities.
565 res = self._search_regex(pattern, string, name, default, fatal, flags, group)
567 return clean_html(res).strip()
571 def _get_login_info(self):
573 Get the login info as (username, password)
574 It will look in the netrc file using the _NETRC_MACHINE value
575 If there's no info available, return (None, None)
577 if self._downloader is None:
582 downloader_params = self._downloader.params
584 # Attempt to use provided username and password or .netrc data
585 if downloader_params.get('username', None) is not None:
586 username = downloader_params['username']
587 password = downloader_params['password']
588 elif downloader_params.get('usenetrc', False):
590 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
595 raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
596 except (IOError, netrc.NetrcParseError) as err:
597 self._downloader.report_warning('parsing .netrc: %s' % compat_str(err))
599 return (username, password)
601 def _get_tfa_info(self):
603 Get the two-factor authentication info
604 TODO - asking the user will be required for sms/phone verify
605 currently just uses the command line option
606 If there's no info available, return None
608 if self._downloader is None:
610 downloader_params = self._downloader.params
612 if downloader_params.get('twofactor', None) is not None:
613 return downloader_params['twofactor']
617 # Helper functions for extracting OpenGraph info
619 def _og_regexes(prop):
620 content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')'
621 property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop)
622 template = r'<meta[^>]+?%s[^>]+?%s'
624 template % (property_re, content_re),
625 template % (content_re, property_re),
628 def _og_search_property(self, prop, html, name=None, **kargs):
630 name = 'OpenGraph %s' % prop
631 escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
634 return unescapeHTML(escaped)
636 def _og_search_thumbnail(self, html, **kargs):
637 return self._og_search_property('image', html, 'thumbnail url', fatal=False, **kargs)
639 def _og_search_description(self, html, **kargs):
640 return self._og_search_property('description', html, fatal=False, **kargs)
642 def _og_search_title(self, html, **kargs):
643 return self._og_search_property('title', html, **kargs)
645 def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
646 regexes = self._og_regexes('video') + self._og_regexes('video:url')
648 regexes = self._og_regexes('video:secure_url') + regexes
649 return self._html_search_regex(regexes, html, name, **kargs)
651 def _og_search_url(self, html, **kargs):
652 return self._og_search_property('url', html, **kargs)
654 def _html_search_meta(self, name, html, display_name=None, fatal=False, **kwargs):
655 if display_name is None:
657 return self._html_search_regex(
659 (?=[^>]+(?:itemprop|name|property)=(["\']?)%s\1)
660 [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(name),
661 html, display_name, fatal=fatal, group='content', **kwargs)
663 def _dc_search_uploader(self, html):
664 return self._html_search_meta('dc.creator', html, 'uploader')
666 def _rta_search(self, html):
667 # See http://www.rtalabel.org/index.php?content=howtofaq#single
668 if re.search(r'(?ix)<meta\s+name="rating"\s+'
669 r' content="RTA-5042-1996-1400-1577-RTA"',
674 def _media_rating_search(self, html):
675 # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
676 rating = self._html_search_meta('rating', html)
688 return RATING_TABLE.get(rating.lower(), None)
690 def _family_friendly_search(self, html):
691 # See http://schema.org/VideoObject
692 family_friendly = self._html_search_meta('isFamilyFriendly', html)
694 if not family_friendly:
703 return RATING_TABLE.get(family_friendly.lower(), None)
705 def _twitter_search_player(self, html):
706 return self._html_search_meta('twitter:player', html,
707 'twitter card player')
710 def _hidden_inputs(html):
712 (input.group('name'), input.group('value')) for input in re.finditer(
715 type=(?P<q_hidden>["\'])hidden(?P=q_hidden)\s+
716 name=(?P<q_name>["\'])(?P<name>.+?)(?P=q_name)\s+
717 (?:id=(?P<q_id>["\']).+?(?P=q_id)\s+)?
718 value=(?P<q_value>["\'])(?P<value>.*?)(?P=q_value)
722 def _form_hidden_inputs(self, form_id, html):
723 form = self._search_regex(
724 r'(?s)<form[^>]+?id=(["\'])%s\1[^>]*>(?P<form>.+?)</form>' % form_id,
725 html, '%s form' % form_id, group='form')
726 return self._hidden_inputs(form)
728 def _sort_formats(self, formats, field_preference=None):
730 raise ExtractorError('No video formats found')
733 # TODO remove the following workaround
734 from ..utils import determine_ext
735 if not f.get('ext') and 'url' in f:
736 f['ext'] = determine_ext(f['url'])
738 if isinstance(field_preference, (list, tuple)):
739 return tuple(f.get(field) if f.get(field) is not None else -1 for field in field_preference)
741 preference = f.get('preference')
742 if preference is None:
743 proto = f.get('protocol')
745 proto = compat_urllib_parse_urlparse(f.get('url', '')).scheme
747 preference = 0 if proto in ['http', 'https'] else -0.1
748 if f.get('ext') in ['f4f', 'f4m']: # Not yet supported
751 if f.get('vcodec') == 'none': # audio only
752 if self._downloader.params.get('prefer_free_formats'):
753 ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
755 ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
758 audio_ext_preference = ORDER.index(f['ext'])
760 audio_ext_preference = -1
762 if self._downloader.params.get('prefer_free_formats'):
763 ORDER = ['flv', 'mp4', 'webm']
765 ORDER = ['webm', 'flv', 'mp4']
767 ext_preference = ORDER.index(f['ext'])
770 audio_ext_preference = 0
774 f.get('language_preference') if f.get('language_preference') is not None else -1,
775 f.get('quality') if f.get('quality') is not None else -1,
776 f.get('tbr') if f.get('tbr') is not None else -1,
777 f.get('filesize') if f.get('filesize') is not None else -1,
778 f.get('vbr') if f.get('vbr') is not None else -1,
779 f.get('height') if f.get('height') is not None else -1,
780 f.get('width') if f.get('width') is not None else -1,
782 f.get('abr') if f.get('abr') is not None else -1,
783 audio_ext_preference,
784 f.get('fps') if f.get('fps') is not None else -1,
785 f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
786 f.get('source_preference') if f.get('source_preference') is not None else -1,
787 f.get('format_id') if f.get('format_id') is not None else '',
789 formats.sort(key=_formats_key)
791 def _check_formats(self, formats, video_id):
794 lambda f: self._is_valid_url(
796 item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
799 def _is_valid_url(self, url, video_id, item='video'):
800 url = self._proto_relative_url(url, scheme='http:')
801 # For now assume non HTTP(S) URLs always valid
802 if not (url.startswith('http://') or url.startswith('https://')):
805 self._request_webpage(url, video_id, 'Checking %s URL' % item)
807 except ExtractorError as e:
808 if isinstance(e.cause, compat_HTTPError):
810 '%s: %s URL is invalid, skipping' % (video_id, item))
814 def http_scheme(self):
815 """ Either "http:" or "https:", depending on the user's preferences """
818 if self._downloader.params.get('prefer_insecure', False)
821 def _proto_relative_url(self, url, scheme=None):
824 if url.startswith('//'):
826 scheme = self.http_scheme()
831 def _sleep(self, timeout, video_id, msg_template=None):
832 if msg_template is None:
833 msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
834 msg = msg_template % {'video_id': video_id, 'timeout': timeout}
838 def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None):
839 manifest = self._download_xml(
840 manifest_url, video_id, 'Downloading f4m manifest',
841 'Unable to download f4m manifest',
842 # Some manifests may be malformed, e.g. prosiebensat1 generated manifests
843 # (see https://github.com/rg3/youtube-dl/issues/6215#issuecomment-121704244)
844 transform_source=lambda s: fix_xml_ampersands(s).strip())
847 manifest_version = '1.0'
848 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
850 manifest_version = '2.0'
851 media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
852 for i, media_el in enumerate(media_nodes):
853 if manifest_version == '2.0':
854 manifest_url = ('/'.join(manifest_url.split('/')[:-1]) + '/' +
855 (media_el.attrib.get('href') or media_el.attrib.get('url')))
856 tbr = int_or_none(media_el.attrib.get('bitrate'))
858 'format_id': '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)])),
862 'width': int_or_none(media_el.attrib.get('width')),
863 'height': int_or_none(media_el.attrib.get('height')),
864 'preference': preference,
866 self._sort_formats(formats)
870 def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
871 entry_protocol='m3u8', preference=None,
872 m3u8_id=None, note=None, errnote=None,
876 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])),
880 'preference': preference - 1 if preference else -1,
881 'resolution': 'multiple',
882 'format_note': 'Quality selection URL',
885 format_url = lambda u: (
887 if re.match(r'^https?://', u)
888 else compat_urlparse.urljoin(m3u8_url, u))
890 m3u8_doc = self._download_webpage(
892 note=note or 'Downloading m3u8 information',
893 errnote=errnote or 'Failed to download m3u8 information',
895 if m3u8_doc is False:
900 r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
901 for line in m3u8_doc.splitlines():
902 if line.startswith('#EXT-X-STREAM-INF:'):
904 for m in kv_rex.finditer(line):
906 if v.startswith('"'):
908 last_info[m.group('key')] = v
909 elif line.startswith('#EXT-X-MEDIA:'):
911 for m in kv_rex.finditer(line):
913 if v.startswith('"'):
915 last_media[m.group('key')] = v
916 elif line.startswith('#') or not line.strip():
919 if last_info is None:
920 formats.append({'url': format_url(line)})
922 tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
925 format_id.append(m3u8_id)
926 last_media_name = last_media.get('NAME') if last_media and last_media.get('TYPE') != 'SUBTITLES' else None
927 format_id.append(last_media_name if last_media_name else '%d' % (tbr if tbr else len(formats)))
929 'format_id': '-'.join(format_id),
930 'url': format_url(line.strip()),
933 'protocol': entry_protocol,
934 'preference': preference,
936 codecs = last_info.get('CODECS')
938 # TODO: looks like video codec is not always necessarily goes first
939 va_codecs = codecs.split(',')
941 f['vcodec'] = va_codecs[0].partition('.')[0]
942 if len(va_codecs) > 1 and va_codecs[1]:
943 f['acodec'] = va_codecs[1].partition('.')[0]
944 resolution = last_info.get('RESOLUTION')
946 width_str, height_str = resolution.split('x')
947 f['width'] = int(width_str)
948 f['height'] = int(height_str)
949 if last_media is not None:
950 f['m3u8_media'] = last_media
954 self._sort_formats(formats)
957 # TODO: improve extraction
958 def _extract_smil_formats(self, smil_url, video_id, fatal=True):
959 smil = self._download_xml(
960 smil_url, video_id, 'Downloading SMIL file',
961 'Unable to download SMIL file', fatal=fatal)
966 base = smil.find('./head/meta').get('base')
970 if smil.findall('./body/seq/video'):
971 video = smil.findall('./body/seq/video')[0]
972 fmts, rtmp_count = self._parse_smil_video(video, video_id, base, rtmp_count)
975 for video in smil.findall('./body/switch/video'):
976 fmts, rtmp_count = self._parse_smil_video(video, video_id, base, rtmp_count)
979 self._sort_formats(formats)
983 def _parse_smil_video(self, video, video_id, base, rtmp_count):
984 src = video.get('src')
986 return ([], rtmp_count)
987 bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000)
988 width = int_or_none(video.get('width'))
989 height = int_or_none(video.get('height'))
990 proto = video.get('proto')
993 if base.startswith('rtmp'):
995 elif base.startswith('http'):
997 ext = video.get('ext')
999 return (self._extract_m3u8_formats(src, video_id, ext), rtmp_count)
1000 elif proto == 'rtmp':
1002 streamer = video.get('streamer') or base
1007 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
1012 elif proto.startswith('http'):
1015 'ext': ext or 'flv',
1021 def _live_title(self, name):
1022 """ Generate the title for a live video """
1023 now = datetime.datetime.now()
1024 now_str = now.strftime("%Y-%m-%d %H:%M")
1025 return name + ' ' + now_str
1027 def _int(self, v, name, fatal=False, **kwargs):
1028 res = int_or_none(v, **kwargs)
1029 if 'get_attr' in kwargs:
1030 print(getattr(v, kwargs['get_attr']))
1032 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1034 raise ExtractorError(msg)
1036 self._downloader.report_warning(msg)
1039 def _float(self, v, name, fatal=False, **kwargs):
1040 res = float_or_none(v, **kwargs)
1042 msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
1044 raise ExtractorError(msg)
1046 self._downloader.report_warning(msg)
1049 def _set_cookie(self, domain, name, value, expire_time=None):
1050 cookie = compat_cookiejar.Cookie(
1051 0, name, value, None, None, domain, None,
1052 None, '/', True, False, expire_time, '', None, None, None)
1053 self._downloader.cookiejar.set_cookie(cookie)
1055 def get_testcases(self, include_onlymatching=False):
1056 t = getattr(self, '_TEST', None)
1058 assert not hasattr(self, '_TESTS'), \
1059 '%s has _TEST and _TESTS' % type(self).__name__
1062 tests = getattr(self, '_TESTS', [])
1064 if not include_onlymatching and t.get('only_matching', False):
1066 t['name'] = type(self).__name__[:-len('IE')]
1069 def is_suitable(self, age_limit):
1070 """ Test whether the extractor is generally suitable for the given
1071 age limit (i.e. pornographic sites are not, all others usually are) """
1073 any_restricted = False
1074 for tc in self.get_testcases(include_onlymatching=False):
1075 if 'playlist' in tc:
1076 tc = tc['playlist'][0]
1077 is_restricted = age_restricted(
1078 tc.get('info_dict', {}).get('age_limit'), age_limit)
1079 if not is_restricted:
1081 any_restricted = any_restricted or is_restricted
1082 return not any_restricted
1084 def extract_subtitles(self, *args, **kwargs):
1085 if (self._downloader.params.get('writesubtitles', False) or
1086 self._downloader.params.get('listsubtitles')):
1087 return self._get_subtitles(*args, **kwargs)
1090 def _get_subtitles(self, *args, **kwargs):
1091 raise NotImplementedError("This method must be implemented by subclasses")
1093 def extract_automatic_captions(self, *args, **kwargs):
1094 if (self._downloader.params.get('writeautomaticsub', False) or
1095 self._downloader.params.get('listsubtitles')):
1096 return self._get_automatic_captions(*args, **kwargs)
1099 def _get_automatic_captions(self, *args, **kwargs):
1100 raise NotImplementedError("This method must be implemented by subclasses")
1103 class SearchInfoExtractor(InfoExtractor):
1105 Base class for paged search queries extractors.
1106 They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query}
1107 Instances should define _SEARCH_KEY and _MAX_RESULTS.
1111 def _make_valid_url(cls):
1112 return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
1115 def suitable(cls, url):
1116 return re.match(cls._make_valid_url(), url) is not None
1118 def _real_extract(self, query):
1119 mobj = re.match(self._make_valid_url(), query)
1121 raise ExtractorError('Invalid search query "%s"' % query)
1123 prefix = mobj.group('prefix')
1124 query = mobj.group('query')
1126 return self._get_n_results(query, 1)
1127 elif prefix == 'all':
1128 return self._get_n_results(query, self._MAX_RESULTS)
1132 raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
1133 elif n > self._MAX_RESULTS:
1134 self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
1135 n = self._MAX_RESULTS
1136 return self._get_n_results(query, n)
1138 def _get_n_results(self, query, n):
1139 """Get a specified number of results for a query"""
1140 raise NotImplementedError("This method must be implemented by subclasses")
1143 def SEARCH_KEY(self):
1144 return self._SEARCH_KEY